To improve performance and reliability of storage systems, redundant array of independent disks (RAID) is widely applied to the storage systems. However, in a large-scale storage system, a disk failure has become a normal state. Therefore, a RAID needs to be reconstructed. Before the RAID is reconstructed, a problem in an aspect of RAID reconstruction triggering needs to be resolved first.
RAID reconstruction triggering solutions currently provided mainly include: a solution 1: triggering reconstruction using dedicated configuration software; and a solution 2: triggering reconstruction in a startup phase of a basic input/output system (BIOS).
However, the solution 1 relies on the dedicated configuration software, and requires a manual operation by a user. In the solution 2, a startup speed is severely affected, the reconstruction takes a long time, and the user easily causes a data loss due to negligence. Therefore, the existing solutions have problems such as low universality and usability.